Regression Models for Forecasting Global Oil Production

Aydin G.

PETROLEUM SCIENCE AND TECHNOLOGY, vol.33, pp.1822-1828, 2015 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 33
  • Publication Date: 2015
  • Doi Number: 10.1080/10916466.2015.1101474
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1822-1828
  • Keywords: oil, production, regression analysis, modeling, forecasting, NATURAL-GAS, ENERGY-CONSUMPTION
  • Karadeniz Technical University Affiliated: Yes


In this study, the global oil production is modeled in order to forecast future projections based on the historical trend using the linear and nonlinear regression analysis. The developed models are validated considering the behavior of determination coefficient, F, and t tests. Forecasting performances of the proposed models are also compared using mean absolute percentage error. The results show that proposed models can be effectively used for forecasting the global oil production. The compared results show that the inverse regression model gives the best forecasting performance. It forecast the global oil production as 4593 Mt in 2020, 8.84% higher than the level in 2014.